INDEX

Explanations

permission and authorization

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æĶ¿åħļ

-0.11

forcing

-0.10

 Forced

-0.10

forced

-0.10

è´´

-0.09

æĤ¬å´ĸ

-0.09

è¯ĺ

-0.09

çº¢æĺŁ

-0.09

çĤ¬

-0.09

 forcing

-0.09

POSITIVE LOGITS

æİĪæĿĥ

0.15

åħģè®¸

0.13

æİĪæ¬Ĭ

0.13

 permission

0.12

 authorize

0.11

æī¹åĩĨ

0.11

æĿĥéĻĲ

0.11

.perm

0.10

åħģ

0.10

è®¸åı¯

0.10

Activations Density 0.072%